Health/Training Front Door

Use this when OpenClaw needs live Fitbit/health/training data in a stable, low-ambiguity way.

This is a thin, typed front door over the canonical Fitbit connector tooling.

Why this exists

fitbit_tools.py is the canonical operational backend, but it still requires low-level CLI composition. This front door provides a narrow action contract so agents can call one stable interface instead of assembling raw shell commands every time.

Contract

Run:

• node skills/health-training-frontdoor/scripts/request.js '{...json...}'

Input JSON:

{
  "action": "latest_recovery"
}

Supported actions:

• auth_status
• latest_recovery
• quality_flags
• training_status
• training_window
• unified_latest

Optional fields:

• days (integer)
• ensureFresh (boolean)
• source (for unified_latest, default best)

Default behavior by action

• latest_recovery: fetches latest days of hrv_rmssd,resting_hr,sleep_minutes,sleep_score,data_quality; defaults days=3, ensureFresh=true
• quality_flags: defaults days=7
• training_status: defaults days=14, ensureFresh=true
• training_window: defaults days=14, ensureFresh=true
• unified_latest: defaults days=14, source=best

Notes

• Output is compact JSON.
• This surface is read-only.
• Interpretation/coaching remains outside this skill.

Training Programming Reference

All programming decisions must be grounded in Practical Programming for Strength Training (Rippetoe & Baker, 3rd Ed.).

Always load before giving any programming advice:

1. reference/practical-programming/INDEX.md — who Joao is, relevant chapters, reading guide
2. memory/training-continuity.md — current lifts, recent sessions, health signals

For any programming question, read the relevant sections of reference/practical-programming/book.md directly. The book is the authority.

Backend mapping

This front door maps directly to:

• fitbit-connector/scripts/fitbit_tools.py

Do not use this to bypass auth or write-capable operations.